Strategic Integration of In Vivo Cardiovascular Models During Lead Optimization: Predictive Value of 4 Models Independent of Species, Route of Administration, and Influence of Anesthesia

Evaluation of the translation of multiple cardiovascular regulatory mechanisms in the anesthetized dog

The strategic and targeted use of an anesthetized canine cardiovascular model early in drug discovery enables a comprehensive cardiovascular and electrophysiological assessment of potential safety liabilities and guides compound selection prior to initiation of chronic toxicological studies. An ideal model would enable exposure-response relationships to guide safety margin calculations, have a low threshold to initiate, and have quick delivery of decision quality data. We have aimed to profile compounds with diverse mechanism of actions (MoAs) of "non-QT" cardiovascular drug effects and evaluate the ability of nonclinical in vivo cardiovascular models to detect clinically reported effects. The hemodynamic effects of 11 drugs (atropine, itraconazole, atenolol, ivabradine, milrinone, enalaprilat, fasudil, amlodipine, prazosin, amiloride, and hydrochlorothiazide) were profiled in an anesthetized dog cardiovascular model. Derived parameters included: heart rate, an index of left ventricular contractility, mean arterial pressure, systemic vascular resistance, and cardiac output. Species specific plasma protein data was generated (human, dog) and utilized to calculate free drug concentrations. Using the anesthetized dog cardiovascular model, 10 of the 11 drugs displayed the predicted changes in CV parameters based on their primary MoAs and corresponding clinically described effects. Interestingly but not unexpected, 1 of 11 failed to display their predicted CV pattern which is likely due to a delay in pharmacodynamic effect that is beyond the duration of the experimental model (hydrochlorothiazide). The analysis from the current study supports the strategic use of the anesthetized dog model early in the drug discovery process for a comprehensive cardiovascular evaluation with good translation to human.

Drug-induced QT prolongation: Concordance of preclinical anesthetized canine model in relation to published clinical observations for ten CiPA drugs

INTRODUCTION

The Comprehensive In Vitro Proarrhythmia Assay (CiPA) initiative differentiates torsadogenic risk of 28 drugs affecting ventricular repolarization based on multiple in vitro human derived ionic currents. However, a standardized prospective assessment of the electrophysiologic effects of these drugs in an integrated in vivo preclinical cardiovascular model is lacking. This study questioned whether QTc interval prolongation in a preclinical in vivo model could detect clinically reported QTc prolongation and assign torsadogenic risk for ten CiPA drugs.

METHODS

An acute intravenous administered ascending dose anesthetized dog cardiovascular model was used to assess QTc prolongation along with other electrocardiographic (PR, QRS intervals) and hemodynamic (heart rate, blood pressures, left ventricular contractility) parameters at plasma concentrations spanning and exceeding clinical exposures. hERG current block potency was characterized using IC₅₀ values from automated patch clamp.

RESULTS

All eight drugs eliciting clinical QTc prolongation also delayed repolarization in anesthetized dogs at plasma concentrations within four-fold clinical exposures. In vitro QTc safety margins (defined based on clinical C_max values/plasma concentrations eliciting statistically significant QTc prolongation in dogs) were lower for high vs intermediate torsadogenic risk drugs. In comparison, hERG IC₁₀ values represented as total drug concentrations were better predictors of preclinical QTc prolongation than hERG IC₅₀ values.

CONCLUSION

There was good concordance for QTc prolongation in the anesthetized dog model and clinical torsadogenic risk assignment. QTc assessment in the anesthetized dog remains a valuable part of a more comprehensive preclinical integrated risk assessment for delayed repolarization and torsadogenic risk as part of a global cardiovascular evaluation.

Cardiovascular safety pharmacology studies in dogs enabled for a poorly soluble molecule using spray-dried dispersion: Impact on lead selection

The aim of this study was to identify an adequate formulation for a poorly soluble lead molecule (BI-A) that would achieve sufficiently high plasma concentrations after oral administration in dogs to enable a robust cardiovascular safety pharmacology assessment in telemetry-instrumented conscious dogs during lead optimization in drug discovery. A spray-dried dispersion of BI-A (BI-A-SDD) containing a 1:2 ratio of BI-A and hydroxypropyl methylcellulose acetate succinate-LF was prepared using a Büchi spray dryer B-90 (B-90). Physical form characterization, an in vitro dissolution test and a preliminary pharmacokinetic (PK) study following oral administration of BI-A-SDD were performed. Thereafter, effects on cardiovascular parameters in conscious, chronically-instrumented dogs were investigated for 24h after a single oral dose (5, 10, and 50mg/kg) using a modified Latin square cross-over study design. The BI-A-SDD powder was confirmed to be amorphous and was stable as an aqueous suspension for at least 4h. The BI-A-SDD suspension provided a greater rate and extent of dissolution than the crystalline BI-A suspension and the supersaturation was maintained for at least 4h. In PK studies the Cmax of the BI-A-SDD formulation (25.4μM; 77-fold the projected efficacious Cmax of 0.33μM) was 7.5-fold higher than the Cmax observed using oral administration of a 10% hydroxypropyl-β-cyclodextrin formulation at 100mg/kg in dogs (3.4μM). In conscious, chronically-instrumented dogs, the doses tested and plasma concentrations achieved were sufficient to enable a robust safety pharmacology evaluation. Multiple off-target hemodynamic effects were detected including acute elevations in aortic blood pressure (up to 22% elevation in systolic and diastolic blood pressure) and tachycardia (68% elevation in heart rate), results that were confirmed in other in vivo models. These results led to a deprioritization of BI-A. The study demonstrated that a spray-dried dispersion, prepared using the B-90 in drug discovery, enhanced the oral exposure of a poorly water-soluble molecule, BI-A, and thereby enabled its evaluation in safety pharmacology studies that ultimately resulted in deprioritization of BI-A from a pool of lead compounds.

Physical Properties and Effect in a Battery of Safety Pharmacology Models for Three Structurally Distinct Enteric Polymers Employed as Spray-Dried Dispersion Carriers

Establishing a wide therapeutic index (TI) for pre-clinical safety is important during lead optimization (LO) in research, prior to clinical development, although is often limited by a molecules physiochemical characteristics. Recent advances in the application of the innovative vibrating mesh spray-drying technology to prepare amorphous solid dispersions may offer an opportunity to achieve high plasma concentrations of poorly soluble NCEs to enable testing and establishment of a wide TI in safety pharmacology studies. While some of the amorphous solid dispersion carriers are generally recognized as safe for clinical use, whether they are sufficiently benign to enable in vivo pharmacology studies has not been sufficiently demonstrated. Thus, the physical properties, and effect in a battery of in vivo safety pharmacology models, were assessed in three classes of polymers employed as spray-dried dispersion carriers. The polymers (HPMC-AS, Eudragit, PVAP) displayed low affinity with acetone/methanol, suitable for solvent-based spray drying. The water sorption of the polymers was moderate, and the degree of hysteresis of HPMC-AS was smaller than Eudragit and PVAP indicating the intermolecular interaction of water-cellulose molecules is weaker than water-acrylate or water-polyvinyl molecules. The polymer particles were well-suspended without aggregation with a mean particle size less than 3 μm in an aqueous vehicle. When tested in conscious Wistar Han rats in safety pharmacology models (n = 6–8/dose/polymer) investigating effects on CNS, gastrointestinal, and cardiovascular function, no liabilities were identified at any dose tested (30–300 mg/kg PO, suspension). In brief, the polymers had no effect in a modified Irwin test that included observational and evoked endpoints related to stereotypies, excitation, sedation, pain/anesthesia, autonomic balance, reflexes, and others. No effect of the polymers on gastric emptying or intestinal transit was observed when measured using a barium sulfate tracer material. Finally, in telemetry-instrumented rats the polymers had no effect on acute or 24-h mean blood pressure and heart rate values at doses up to 300 mg/kg. Thus, the properties of the three enteric polymers are appropriate as spray-dried dispersion carriers and were benign in a battery of safety pharmacology studies, demonstrating their applicability to enable in vivo safety pharmacology profiling of poorly soluble molecules during LO.

G protein-coupled bile acid receptor 1 stimulation mediates arterial vasodilation through a K(Ca)1.1 (BK(Ca))-dependent mechanism

Bile acids (BAs) and BA receptors, including G protein-coupled bile acid receptor 1 (GPBAR1), represent novel targets for the treatment of metabolic and inflammatory disorders. However, BAs elicit myriad effects on cardiovascular function, although this has not been specifically ascribed to GPBAR1. This study was designed to test whether stimulation of GPBAR1 elicits effects on cardiovascular function that are mechanism based that can be identified in acute ex vivo and in vivo cardiovascular models, to delineate whether effects were due to pathways known to be modulated by BAs, and to establish whether a therapeutic window between in vivo cardiovascular liabilities and on-target efficacy could be defined. The results demonstrated that the infusion of three structurally diverse and selective GPBAR1 agonists produced marked reductions in vascular tone and blood pressure in dog, but not in rat, as well as reflex tachycardia and a positive inotropic response, effects that manifested in an enhanced cardiac output. Changes in cardiovascular function were unrelated to modulation of the levothyroxine/thyroxine axis and were nitric oxide independent. A direct effect on vascular tone was confirmed in dog isolated vascular rings, whereby concentration-dependent decreases in tension that were tightly correlated with reductions in vascular tone observed in vivo and were blocked by iberiotoxin. Compound concentrations in which cardiovascular effects occurred, both ex vivo and in vivo, could not be separated from those necessary for modulation of GPBAR1-mediated efficacy, resulting in project termination. These results are the first to clearly demonstrate direct and potent peripheral arterial vasodilation due to GPBAR1 stimulation in vivo through activation of large conductance Ca(2+) activated potassium channel K(Ca)1.1.

The clinically-tested S1P receptor agonists, FTY720 and BAF312, demonstrate subtype-specific bradycardia (S1P₁) and hypertension (S1P₃) in rat

Sphingosine-1-phospate (S1P) and S1P receptor agonists elicit mechanism-based effects on cardiovascular function in vivo. Indeed, FTY720 (non-selective S1P(X) receptor agonist) produces modest hypertension in patients (2-3 mmHg in 1-yr trial) as well as acute bradycardia independent of changes in blood pressure. However, the precise receptor subtypes responsible is controversial, likely dependent upon the cardiovascular response in question (e.g. bradycardia, hypertension), and perhaps even species-dependent since functional differences in rodent, rabbit, and human have been suggested. Thus, we characterized the S1P receptor subtype specificity for each compound in vitro and, in vivo, the cardiovascular effects of FTY720 and the more selective S1P₁,₅ agonist, BAF312, were tested during acute i.v. infusion in anesthetized rats and after oral administration for 10 days in telemetry-instrumented conscious rats. Acute i.v. infusion of FTY720 (0.1, 0.3, 1.0 mg/kg/20 min) or BAF312 (0.5, 1.5, 5.0 mg/kg/20 min) elicited acute bradycardia in anesthetized rats demonstrating an S1P₁ mediated mechanism-of-action. However, while FTY720 (0.5, 1.5, 5.0 mg/kg/d) elicited dose-dependent hypertension after multiple days of oral administration in rat at clinically relevant plasma concentrations (24-hr mean blood pressure = 8.4, 12.8, 16.2 mmHg above baseline vs. 3 mmHg in vehicle controls), BAF312 (0.3, 3.0, 30.0 mg/kg/d) had no significant effect on blood pressure at any dose tested suggesting that hypertension produced by FTY720 is mediated S1P₃ receptors. In summary, in vitro selectivity results in combination with studies performed in anesthetized and conscious rats administered two clinically tested S1P agonists, FTY720 or BAF312, suggest that S1P₁ receptors mediate bradycardia while hypertension is mediated by S1P₃ receptor activation.